Peptides: What They Are, and Why the Rules Keep Changing

Peptides occupy a strange place in medicine right now. They are simultaneously oversold by people who should know better and dismissed by people who have not looked. Both reactions miss what is actually going on.

Here is what I tell patients who ask.

What a peptide actually is

A peptide is a short chain of amino acids. Proteins are long chains of the same building blocks. The distinction is mostly length.

This matters because your body already runs on peptides. Insulin is one. So is oxytocin, and so is the growth hormone releasing hormone your hypothalamus produces every night. These are not exotic substances. They are the signaling molecules your endocrine system already uses to talk to itself, and several of them are already sitting in ordinary pharmacies as ordinary medications.

Most therapeutic peptides work by imitating one of those signals. They bind a receptor your body already has and trigger a response your body already knows how to make. That is the mechanistic argument for them, and it is a reasonable one.

It is not, by itself, evidence that any particular peptide will do anything useful for you.

Why so few of them are FDA approved

This is the part almost nobody explains, and it is the key to understanding the whole field.

Only a handful of peptides have gone through full FDA approval. Semaglutide and tirzepatide, the GLP-1 medications. Bremelanotide. Tesamorelin. Oxytocin. That is close to the complete list.

FDA approval requires multi-phase human trials that cost hundreds of millions of dollars. That money comes from pharmaceutical companies, and companies spend it when they hold a patent that lets them recoup it.

Many therapeutic peptides are naturally occurring molecules. You cannot patent a molecule the human body already makes. No patent means no exclusivity, no exclusivity means no way to recover a nine-figure trial cost, and so nobody funds the trials.

The result is a genuine regulatory gap. Some compounds have decades of research behind them and remain unapproved in the United States, not because anyone found them unsafe, but because no company had a financial reason to pay for the paperwork.

That is worth sitting with, because it cuts both ways. "Not FDA approved" does not mean unsafe or ineffective. It also does not mean proven. It frequently means nobody has done the work to find out, and you should want to know which situation you are in.

How the compounding rules work

Because of that gap, most peptides reach patients through compounding pharmacies rather than as commercial products. The FDA maintains lists governing what pharmacies may compound.

Category 1 substances may be compounded. Category 2 substances may not, having been evaluated and restricted. In between sits a group under active review by an FDA advisory committee, which examines the evidence on each compound and recommends whether it moves back to Category 1 or stays restricted.

That middle group is where most of the buzz lives, and it is the part I would be most careful about. In 2026 a batch of well-known peptides was removed from the restricted list but not approved for compounding, pending review. Regulatory limbo is not the same as clearance. Any pharmacy or clinic selling a compound while it sits in that gap is operating in a gray area, and you are the one carrying the risk.

The five questions I ask before any peptide

Before I will prescribe a peptide, or take one myself, it has to survive the same five questions. They are not original to me. This is the standard framework of drug evaluation, and it applies to a compounded peptide exactly as much as to anything a pharmaceutical company sells.

1. Is there a believable mechanism? Not a story about one, an actual pathway we can point to. 2. Has it shown a real benefit in humans? Placebo-controlled studies in people, not mouse data and not testimonials. This is where most of the excitement dies. 3. Do we understand the safety? Dosing, how it moves through the body, what the long-term risks are, and what we would need to monitor while you are on it. For many peptides the honest answer is that nobody knows, because nobody has looked for long enough. 4. Does the benefit justify the risk? A modest, uncertain upside does not earn an unknown downside. 5. Is there a better-characterized way to get the same result? This one kills more peptides than any other. If a well-studied drug, or sleep, or strength training gets you the same place, the exotic option has to beat it, not just match it.

Peter Attia, whose podcast has covered this territory more carefully than most, sorts peptides into three buckets, and I find the framing useful whatever you think of him. Roughly: a first bucket that is scientifically unsupported, with no validated mechanism and no credible human evidence. A second where the biology may be real but the claimed clinical benefit has not been demonstrated, or where the harms that surfaced are the kind that would make a company developing the molecule walk away from it. And a third, the smallest, where the sensible move is to use the best-characterized product available, with real oversight. Most of what is marketed as peptide therapy lives in the first two buckets while borrowing the reputation of the third.

I am deliberately not publishing a list of what is available today, because that list has changed repeatedly and will change again. If someone offers you a peptide, the useful question is not whether it sounds impressive. It is what its current regulatory status is, and whether they can tell you without hedging.

Some of the ones I use

Peptides are not a fringe category, and you may already be taking one without thinking of it that way.

Insulin is a peptide, and it has been in clinical use for a century. Semaglutide and tirzepatide, the GLP-1 medications behind the entire current weight loss conversation, are peptides. Bremelanotide, used for low sexual desire, is a peptide. So is oxytocin. I prescribe several of these routinely.

None of them get marketed as peptide therapy, and that is the tell. Once a compound clears FDA approval it stops being called a peptide and simply becomes a drug. The word tends to survive only where the evidence has not caught up yet.

So when someone pitches you peptides as a category, notice that the best-supported ones have quietly left the category.

Sermorelin is the compounded one I am most comfortable with, and the reasons are instructive.

It is a synthetic version of the first portion of your own growth hormone releasing hormone. It binds receptors on the pituitary and prompts your pituitary to release growth hormone in its own natural rhythm.

That upstream position is the whole point. Because sermorelin works on the pituitary rather than replacing growth hormone directly, the release stays under your body's own feedback control. When growth hormone and IGF-1 rise, the system dampens itself, the way it is supposed to. Injecting growth hormone directly bypasses that safety mechanism entirely.

Sermorelin was FDA approved in 1990 for pediatric growth hormone deficiency, sold as Geref. It was withdrawn from the market in 2008 for business reasons rather than safety ones, and it has remained legally compoundable ever since.

What it is reasonably used for in adults is modest: improvement in lean body mass, reduction in visceral fat, better sleep quality, and better recovery. It is particularly useful for preserving muscle in people losing weight on GLP-1 medications, where losing muscle alongside fat is a real and underappreciated problem.

What it is not is an anti-aging drug, and anyone selling it as one has moved past what the evidence supports.

What monitoring should look like

If you are on a growth hormone axis peptide and nobody is checking your IGF-1, that is a problem.

IGF-1 is the downstream marker that tells us whether the growth hormone axis is where it should be. The goal is the upper end of the normal range for your age and sex, not above it. Driving it higher is not better, and sustained elevation is where the theoretical risks of this class live: worsening insulin resistance, and the concern that pushing growth signaling could accelerate a cancer nobody knew was there.

That last point is why active malignancy is a firm contraindication, and why this is not something to buy online and self-administer.

Reasonable monitoring means IGF-1 and metabolic labs before starting, repeat testing after a course of treatment, and a real decision point based on the result. If IGF-1 comes back elevated, the answer is to stop and let it normalize rather than push on.

How to evaluate anything you are offered

Ask where the compound sits on the evidence spectrum. FDA approved means full human trials. Strong clinical evidence means real human data, sometimes smaller studies. Moderate means a mix of small human studies and good animal work. Early or preclinical means the mechanism makes sense and the human data is thin or absent.

Most of the peptides generating excitement online are in that last group. That does not make them worthless. It makes them experiments, and you should be told you are in one.

Ask about the pharmacy too. Compounding pharmacies vary enormously. A good one will provide a certificate of analysis showing purity and identity testing on the batch you are getting. If nobody can produce one, you do not actually know what is in the vial.

Where I land

I use peptides selectively and I am comfortable saying which ones and why. Sermorelin has a defensible mechanism, a long safety record, and honest, modest expectations attached to it. As the review process resolves, other compounds may earn the same footing, and I will use them when they do.

I am not trying to talk anyone out of the category, and I am glad to work with patients who arrive informed and want to make intelligent decisions about it. My job is to make sure anything you use has a real mechanism, a defensible safety picture, clear regulatory standing, and monitoring while you are on it.

What I will not do is sell you a compound in regulatory limbo on the strength of animal studies and a good story. The peptide field has a real problem with people doing exactly that, and the compounds that deserve serious attention get tarred by it.

If you want to know whether a specific peptide makes sense for you, bring me the name and I will tell you honestly what we know, what we do not, and whether it is even legal to prescribe this month.